The Reactor Antineutrino Anomaly (RAA) was highlighted in 2011, when a reevaluated prediction of reactor antineutrinos spectra showed a 6% deficit in the rates observed by previous reactor experiments. A possible explanation for this anomaly consists in introducing a sterile neutrino state at the eV mass scale, participating to the neutrino oscillations mecanism only. The STEREO experiment has been designed to probe the phase-space region indicated by the RAA, by placing a segmented antineutrino target at 10m from the virtually pure 235-uranium compact core of the ILL research reactor facility, in Grenoble. Antineutrinos are detected via inverse beta decay (IBD) in gadolinium-doped liquid scintillator. An oscillation pattern - if any - would develop along the six identical target cells. The relative comparison of their energy distributions allows to test the sterile neutrino hypothesis with reduced systematic uncertainties and without referring to an external prediction. This work focuses mainly on data analysis. The IBD signature is a two-fold signal, requiring a proper pair search algorithm for extracting the antineutrino candidates. After selections, the residual cosmogenic correlated background in the region of interest comes from muon decays and neutron induced reactions. A modelisation of the Pulse Shape Discrimination (PSD) distributions of this background is used to extract the antineutrino signal. Based on the resulting spectra, the first STEREO oscillation analysis rejects the RAA best fit point at 98% of confidence level.